Show HN: InfiniDB – The Unreliable Database of Everything (tncardoso.com)

InfiniDB has emerged as an experimental project that reimagines how large language models (LLMs) function as databases. Created during the holiday season, InfiniDB allows users to dynamically generate tables and query them using familiar SQL commands. Built on SQLite, it utilizes a virtual table module that calls upon LLMs to create table schemas and populate them with data upon the first query. This innovative approach enables users to interact with various datasets, such as a list of Beatles members or the original 151 Pokémon, showcasing the versatility of LLMs in handling diverse information. Significantly, InfiniDB highlights a novel intersection of AI and database technology, opening avenues for further exploration within the AI/ML community. While providing a fun and engaging way to access information, it also raises questions around reliability, as both the schema and data can vary unpredictably due to LLM limitations. Developers interested in the interplay between artificial intelligence and data management can explore the project's potential, though it is advised that users exercise caution, as InfiniDB is currently not suited for production-level applications. The source code is available on GitHub for those wishing to experiment further.